There is a new bioinformatics contest: The Assemblathon. The idea is to provide a CASP-like competition for judging the quality of genome assembly methods from next-generation sequence data.
Actually there are two contests, the Assemblathon and the de Novo Genome Assessment Project (dnGASP), which is based in Europe.
I’ve been thinking that I ought to be entering the Assemblathon, since genome assembly is one of the fields I’m planning to investigate for my sabbatical, but I have no software for assembling Illumina reads (or simulations of them, which is what the Assemblathon is using). The deadline for submitting assemblies is Feb 6, so there is certainly no time for developing new software (even if I had time for research with my teaching load this year). My main interest is in small genomes (prokaryotes), but the Assemblathon is focused on diploid genomes about the size of a human chromosome.
Of course, the results of the Assemblathon will be coming out just in time for the Banana Slug Genomics class this spring, and we’ll certainly be looking at the results very carefully and trying to get our hands on the software that performs best, because the banana slug genome is probably at least as difficult as the Assemblathon pseudo-genome. (It may be polyploid and it is almost certainly bigger than one human chromosome.)
Gas station without pumps 
Hi,
We hope that, if we secure funding, we can run subsequent Assemblathons in the future and broaden the contest into assembly of metagenomes & transcriptomes, use different styles of next-gen data etc. There will definitely be an Assemblathon 2 later this year with real-world data (which also means that we won’t know what the actual answer is). We would also like to take what we have learnt about simulating genomes to create a larger, and more realistic synthetic genome.
Regards,
Keith
Comment by Keith Bradnam — 2011 February 7 @ 13:23 |